Age, sex, and breed evaluations of 392 dogs with chronic valve disease (CVD) from a survey of 4,831 dogs revealed a predominance of CVD in purebred male dogs. Male Cocker Spaniels were most frequently affected. A questionnaire survey concerning aspects of medical history, environment, behavior, and survival was made 3-5 years after initial clinical examinations on 471 dogs. This study revealed no marked differences between dogs with CVD and dogs with no heart disease, except that dogs with CVD more often had signs of congestive heart failure. Endocardial and endomyocardial splitting of the left atrium occurred in 30 dogs with CVD; primarily in males of the Dachshund and Cocker Spaniel breeds. Left atrial perforation in several of the cases caused hemopericardium or acquired atrial septal defects which could be diagnosed by clinical means. The cause of splitting was considered to be left atrial dilatation with left atrial degeneration playing primarily a permissive role. Lipid deposition in the endocardium and ruptured chordae tendineae was found in preliminary frozen section studies. A method of postmortem cardiotomy was developed which permitted recognition of spontaneously ruptured cordae tendineae were frequently found in dogs with severe chronic valve disease with or without left atrial splitting. The Dachshund, Cocker Spaniel, and Beagle breeds in varying order constituted the three most common breeds with CVD, left atrial splitting or the intervertebral disc syndrome. This finding was suggestive of an underlying connective tissue disorder predominant in those breeds which have been classified by others as belonging to a "chondrodystrophoid" group. Radiographic studies of dogs with left atrial enlargement revealed a characteristic bulge in the dorsoventral cardiac silhouette caused by dilatation of the left atrial appendage. This was most apparent in dogs with left atrial splitting. Clinical and pathological studies on 60 "hyperimmune" dogs from which blood was withdrawn by cardiocentesis for immune serum production revealed serious heart disease in nearly all dogs. This was considered the results of weekly cardiocentesis for periods up to three years rather than the results of repeated vaccinations during this time.